Mario Giulianelli

@glnmario

Associate Prof @ucl, Co-founder @parallx_ai | Member of @ELLISforEurope | | Prev. senior research scientist @AISafetyInst, postdoc @ETH_en, PhD @illc_amsterdam

glnmario.github.io

Joined October 2011

Replies Reposts Media

Pinned
Mario Giulianelli
@glnmario
Mar 25
We introduce incremental alternative sampling (IAS), a computational account of language comprehension in which the comprehender continually generates, evaluates, and updates predictions about alternative continuations of partial linguistic input. [1/N]
1
Mario Giulianelli
@glnmario
Sep 9
Join our founding team to help us shape the future of frontier AI systems audits!
Parallax
@parallx_ai
Sep 9
Introducing Parallax 🔍 We're a new UK/EU-based nonprofit research lab building scalable methods and infrastructure to audit the beliefs, goals, and plans that shape model behaviours in long-horizon agentic evaluations. parallx.ai — Thread 🧵 1/
Mario Giulianelli
@glnmario
Aug 20
I've noticed that LLMs are very good at interpreting what students and sometimes collaborators mean in project updates (variables undefined, terminology unexplained, etc.) - I wonder if this is remarkable mind reading, or simply due to the texts being written themselves by an LLM
1
Mario Giulianelli
@glnmario
Jul 20
I really don't see how one could possibly write good meta reviews for 8 papers in 4 days, where each paper is 8 pages + at least as many pages of reviews and rebuttals (what happened to rebuttal length limits btw). how do people do this realistically?
7
Mario Giulianelli
@glnmario
Jul 15
Only a few days left to submit your paper to the LM Playschool Workshop #EMNLP2026! We welcome work on LMs that learn, adapt, or improve through situated, conversational, collaborative, task-oriented, or multi-turn interaction. Deadline: 17 July AoE
lm-playschool.github.io
The LM Playschool Workshop
A Playschool for LLMs
2